Sampled normals are zero

[dipaco]

Hi,

Thank you some much for sharing this project. I am working on similar projects, and this is very useful. I am actually trying to use the sampled normal you provide here. However, on some meshes the normals appear to be zero. Is this an error, or I am loading the normals in the wrong way. I am simply using np.load.

Thanks

[KunalMGupta]

Hi dipaco,

Simply loading with np.load should work. I don't think there should be zeros in sampled normals. Can you point me to specific models that you think have this issue? I can look into it then.

Thanks,

Kunal

[dipaco]

Hi Kunal,

Thanks for replying. I am loading the data using the following script:

In [1]: import numpy as np
In [2]: normals = np.load('normals.npy')
In [3]: normals.shape
Out[3]: (188864, 3)
In [4]: np.linalg.norm(normals, axis=1)
Out[4]: array([0., 0., 0., ..., 0., 0., 0.])
In [5]: np.linalg.norm(normals, axis=1).min()
Out[5]: 0.0
In [6]: np.linalg.norm(normals, axis=1).max()
Out[6]: 0.0

This happened with the model: 02958343/fffb1660a38af30ba4cf3601fb6b2442 last time I tried. But hapens with others as well. I am using python 3.6, and downloaded the data from the provided link.

Thank you.